The medical records of patients with tuberculosis hospitalized in 1966 and 1967 were investigated retrospectively for the incidence of SMON and the intake of chinoform and other drugs during the same calendar years. Those who were given chinoform for the treatment of abdominal symptoms such as pain and diarrhea lasting for at least a day showed a significantly higher incidence of SMON than those who had similar abdminal symptoms but were treated with other drugs. When the comparison was restricted to those who had had the abdominal symptoms for at least a week, a similar higher incidence of SMON was noted for those treated with chinoform than for those treated with other drugs, but the difference was not statistically significant. It was also demonstrated that five SMON patients whom we observed had been given larger amounts of chinoform than those who had been given chinoform but did not develop SMON. It was remarkable that one patient who had been given more than 200g of chinoform from January to October, 1966, did not develop SMON. None of the 12 drugs used for the treatment of tuberculosis seemed to be associated with the disease.